Search

Aleutian Tern Onychoprion aleuticus numbers in Alaska appear to be in decline; however, colonies are difficult to count for the purposes of monitoring due to their co-occurrence with Arctic Terns Sterna paradisaea, low nesting densities, high variability in attendance, sensitivity to human disturbance, and remote terrain. We paired visual observations with unmanned aerial systems (UASs) to test the feasibility of using this technology to survey a small colony of nesting Arctic and Aleutian terns in southcentral Alaska. We used counts of terns in the air and nest attendance to test for disturbance from UAS flights. We conducted 11 UAS flights over seven days at altitudes ranging from 15-30 m and located 23 nests in the 4.1 ha (0.041 km²) colony site (5.6 nests ha-1) by systematically searching the orthomosaics. We were most likely to distinguish tern species in the 18-m and 15-m altitude photos; the white forehead was most visible in side angle rather than top-down. Nest attendance and the number of aerial birds were not influenced by the UAS in this colony, which experiences high Larus gull activity. Aleutian Terns arrived on 19 May and were attending nests by 02 June; nest attendance declined throughout June. We believe the best approximation of total nesting pairs will be achieved by counting nests 7-10 d after first initiation. In mixed species colonies, we recommend that UASs fly in a 'lawn mower' pattern with overlap to maximize side angle images and aim for an image resolution of < 4 mm (which was achieved by 15-m altitude flights with our camera setup). In single species colonies, 30 m is sufficient and can be achieved more efficiently. [ABSTRACT FROM AUTHOR]

The golden-cheeked warbler (Dendroica chrysoparia) is a federally listed endangered species with an extremely limited breeding range (Edwards Plateau, Tex., USA) and specific habitat requirements. Although localized habitat composition and structure correlate well with golden-cheeked warbler occurrence and nest location, the effects of landscape composition and structure are largely untested. TO investigate landscape-scale influences on habitat occupancy, we surveyed 202 point locations for golden-cheeked warbler presence. We used logistic regression to compare the relative influence of 4 landscape-scale variables (largest patch index, percent of the landscape in woodland, mean-nearest neighbor, and edge density) measured at 4 spatial scales (3 ha, 12 ha, 50 ha, and 200 ha) in predicting occurrence. The amount of mature juniper-oak (Juniperus ashei and Quercus spp.) woodland in the landscape was the single most important variable across spatial scales and average values of woodland cover remained stable across scales. Golden cheeked warblers occurred in a local habitat only when landscape composition exceeded 40% woodland and the likelihood of occurrence was greater than 0.5 only when landscape composition exceeded 80% woodland, By contrast, common measures of habitat fragmentation, including edge density and mean-nearest neighbor, were poor predictors of species occurrence across all spatial scales. Our results suggest that golden-cheeked warblers may be less sensitive to fragmentation then they are to overall habitat loss, which may be due to natural habitat heterogeneity related to the variable terrain. Our results do not preclude a relationship between habitat fragmentation and overall habitat quality as measured by nesting success, survival, and recruitment. While our results hold promise for increasing the reliability of predicting species occurrence, future studies are needed to determine the influence of landscape composition and structure on the ultimate measures of habitat quality. [ABSTRACT FROM AUTHOR]

Spruce beetle-induced (Dendroctonus rufipennis (Kirby)) mortality on the Kenai Peninsula has been hypothesized by local ecologists to result in the conversion of forest to grassland and subsequent increased fire danger. This hypothesis stands in contrast to empirical studies in the continental US which suggested that beetle mortality has only a negligible effect on fire danger. In response, we conducted a study using Landsat data and modeling techniques to map land cover change in the Kenai Peninsula and to integrate change maps with other geospatial data to predictively map fire danger for the same region. We collected Landsat imagery to map land cover change at roughly five-year intervals following a severe, mid-1990s beetle infestation to the present. Land cover classification was performed at each time step and used to quantify grassland encroachment patterns over time. The maps of land cover change along with digital elevation models (DEMs), temperature, and historical fire data were used to map and assess wildfire danger across the study area. Results indicate the highest wildfire danger tended to occur in herbaceous and black spruce land cover types, suggesting that the relationship between spruce beetle damage and wildfire danger in costal Alaskan forested ecosystems differs from the relationship between the two in the forests of the coterminous United States. These change detection analyses and fire danger predictions provide the Kenai National Wildlife Refuge (KENWR) ecologists and other forest managers a better understanding of the extent and magnitude of grassland conversion and subsequent change in fire danger following the 1990s spruce beetle outbreak. [ABSTRACT FROM AUTHOR]